Rankings out yesterday, I think the teams that deserve to be 1-6 were there. All first place votes went to BA, which shows that the ranking committee is looking at total body of work from the start of the season. Farmingdale is a solid second with a very nice second ranking period. I looked at both schedules from here on out after the Christmas break, and it looks like the MOST each team will have is maybe two more losses, if that. So I agree with Lucky, 1 and 2 are pretty much set unless a team experiences a complete breakdown or numerous injuries occur (which no one hopes happens).
Cal received a fair amount of votes and is a solid 3, Rowan, Southern and Fairfield are respectively 4, 5 and 6. Voting was very close here. I think any of these three teams can beat anyone on a given night, which Fairfield did the last game before the break. Anyone have any opinion on the rankings?
I’d agree with you and fries , however Fairfield beat Cal and Farmingdale so I don’t agree with them being the 6 at all…
Rider, I don’t think Fairfield beat Rowan, I believe Rowan beat them 2-1 in their only matchup.
You are right but I don’t agree with them being the 6
For the record… This is why we are going to computer rankings! Yes the correct teams are there, but holy cow 3-6 is out of order. I know it’s hard to quantify who’s better then who, when you don’t see everyone play, or don’t play common schedules, but this is as bad a spread of votes I’ve seen in quite some time in my humble opinion.
I’m going to get killed for this but… CAL should not be in the top 4. Last year we jumped all over Neumann for their huge win over Farmingdale early in the year, watched them blow lesser teams out and it wasn’t until February that they fell from the top 2. This year’s CAL team is a very good team, but other then a split with Farmingdale and an OT loss to BA, what have they done. Fairfield beat them handily, split with Farmingdale and lost to BA in OT also, and has played a much harder schedule to date (The Empire is a stronger conference then the CHE this year). So to have Southern over Fairfield and both under Rowan and CAL is based on what?
Somebody please attempt to justify that for me! With logic and tangible data… Because this is how I’m looking at this.
Rowan, Fairfield, CAL and Southern are clearly 3-6.
CAL - lost to Fairfield, doesn’t paly Rowan, doesn’t play Southern
Rowan - beat Fairfield (1 more to play), doesn’t play CAL, doesn’t play Southern
Southern - split with Fairfield (1 more to play), doesn’t play CAL, doesn’t play Rowan
Fairfield - beat CAL, split with Southern (1 more to play), lost 2-1 to Rowan (1 more to play)
This is why the computer is coming. Nothing personal, just reality! Too many agenda’s!
I agree completely fries. Hope to see things sort themselves out by final rankings. Is the computer rankings a definite for the future? Could be wrong but D2 just made the jump right?
To my knowledge yes. I know there was a lot of discussion at Naples last year around the pool about it. D2 is using it. I’ve been looking at the computer rankings a great deal, and think there are a few little things they have to tweak, but they are pretty accurate. The best part is that they are comprehensive, so you see the impact of a weekend every Wednesday, and by January the accuracy of the data is very high. I’m curious to see how the computer sees the Atlantic in the final ranking, because the disparity between the computer top 15 and the human top 15 is pretty funny. We have teams ranked in the top 10 that are out of the top 15 on the computer side.
Gents,
Usually steer clear of the ranking debate, but you collectively have asked some very pertinent questions to the ACHA and M3 with respect to rankings and the post season.
I’ll make every attempt to just keep to the facts. Opinions and feedback, of course, are always welcomed.
- M1, M2, W1, and W2 are all using variants of computer rankings.
A) M1 has an algorithm that weighs W-L% stronger than SoS.
B) M2 has an algorithm that gives SoS the most weight (growing stronger as the season progresses). The final rating
Gents,
Usually steer clear of the ranking debate, but you collectively have asked some very pertinent questions to the ACHA and M3 with respect to rankings and the post season.
I’ll make every attempt to just keep to the facts. Opinions and feedback, of course, are always welcomed.
- M1, M2, W1, and W2 are all using variants of computer rankings.
- M1 has an algorithm that weighs W-L% stronger than SoS.
- M2 has an algorithm that gives SoS the most weight (growing stronger as the season progresses).
- M2 Final Rating = SoS + AGD + W-L%. AGD and W-L% have much less weight than SoS
- M3 Computer Rankings are being tracked this year for evaluation using M2 algorithm. We look at the M2/M3 rankings each week - the preliminary result is that the AGD addition might hold too much weight. Will keep evaluating.
- M3 will have computer rankings on the agenda in Naples for vote. Obviously, with all other divisions using computer rankings, M3 is last to embrace the technology.
- M3 Atlantic Computer Rankings are currently:
#1 Bryn Athn
#2 Farmingdale
#3 Fairfield
#4 S CT
#5 Rowan
#6 Cal UPenn
#7 Fredonia
#8 PSU-Altoona (weakest SoS in top 10, but highest AGD)
#9 Quinnipiac
#10 GMU - M1 Rankings have been in effect for several years with fairly consistant results:
A) Extemely low (<15% upset ratio in National Tournament)
B) #1-#3 as National Champions
C) #9 ISU lowest seed to make National Finals last year - prior to that #6. - M2 Computer Rankings have a few anomalies, but we constantly examine why the teams are ranked where they are - it’s competely objective. As Fries stated, we need to continue to assess and tweak. Tweaks thus far:
A) Elimination of non-ACHA opponent games.
B) Elimination of counting forfeits (whether the score is entered 1-0 or 7-0, can adversely affect the computer rankings (i.e. CFU 1-0(F)'s over College of Charleston actually lower CFU rating while increasing CoC).
C) Under examination:
a) Half the AGD addition
b) Give teams that forfeit, are late with payments, or are late with eligibility forms, etc, a -1 on permanent rating/incident. - Very strong majority of members in M1 & M2 prefer the weekly computer rankings over the previous human ranking committees.
- Much of the stress comes from the creation of the #1 and #2 National Autobids - the Board and Commissioners are working hard to develop a more expansive, but all inclusive Regionals (no autobids). The current system places far too much of a premium on being #2, when in fact, the difference between the top 3 teams in each region is negligible, M2 or M3.
Always open for feedback to make the ACHA a better organization for it’s entire body of membership. In the end, M3 will determine which ranking system will be used in the future. Will place in survey monkey next month to M3 members.
S
Solid post, thanks sheriff
while I like the computer rankings and think they should be used. Seeing PSU-Altoona is the top 10 is the reason why they will need some significant work. They have barely beat any actual teams. They beat up on a ton of below cupcakes.
adq, as always appreciate the feedback. However, there are several quick points I would like to make on your assessment:
- There is a quick fix to the algorithm that will correct the perceived over assessment of PSU-Altoona (and others that run up the score on weaker opponents), should the committee deem necessary. You have identified a minor issue in the computer algoritm that places perhaps too much weight on running up the score on weaker opponents, which others have identified as well. Have already worked this solution into next year’s computer rankings for M2 (dropping teams in rating that have extraordinary high AGD’s). However, this, in my humble opinion, does not qualify as “significant work” - rather simple fix, really.
- The M3 debate always appears to be a one-sided debate pointing out a singular perceived flaw among 40 teams (the top ten in each region) - in this case #8 in the Atlantic, #45 in the Nation - halfway through the season - really? The debate should be, again, in my humble opinion, does the computer get it right more than the human rankers in the top 20 Nationally - past history in the entirety of the ACHA (M1, M2, W1/W2 all use computer rankings) would suggest this is certainly the case.
- Thanks to your feedback, the feedback by coaches, regional commissioners, commissioners, board members, and this blog, we have made a few minor adjustments to next year’s algorithm. We have run these changes behind the scenes on this season to examine the alterations to the ratings on various teams and believe the rankings will improve as a result.
- As I clearly stated in previous blogs, email, and texts - having been involved for several years in both, first as a human ranker, then in studying possible algorithms, I lean heavily to the computer. Besides being far less biased than the human board, the computer rankings take far less time and effort, and can be displayed every week after 1 Nov. It also eliminates all the back room chatter with the ranking committee that is never admitted to, but almost always occured (in my experience), that inevitably affects the final rankings of the final ranking period.
- Having said all that, this is clearly an M3 decision. If/when the membership votes to incorporate the computer rankings, the ACHA will be ready to place in motion. Until then, the current system has worked adequately for most members for many seasons. There is no perfect ranking system, only constant improvement and persistant pursuit of perfection to ultimately benefit the student-athletes of the ACHA.
I look forward to continued dialogue on this subject - feedback always leads to process improvement.
PSU-Altoona drops from #8 to #10 this week given a good loss to PSU and a bad loss to SCSU…not a bad top 10 this week. Will continue to refine as the season progresses.